Utilizing the Medication Event Monitoring System (MEMS; Aprex Corp.), a specialized pill bottle cap to record time and date of pill bottle opening, this study will determine whether two adherence-focused counseling interventions of graded intensity enhance MEMS-assessed adherence to antiretroviral medication compared to a control intervention in HIV-positive patients. The two adherence interventions are Review of MEMS displayed adherence data - ROM, and Contingency-added to review of MEMS-displayed adherence data - CAROM. Patients currently prescribed anti-retroviral medication will be enrolled. The principle outcome measure in this study will be rate of MEMS cap opening within the appropriate time frame. Secondary measures of adherence will include MEMS-assessed adherene to correct number of pills taken and to number of days a medication was taken; pill counts; patient estimates of adherence; attendance at scheduled study appointments; means corpuscular volume; and plasma levels of anti-retrovirals (where applicable).